CIO interview: NHS IT chief Christine Connelly on the National Programme for IT

As director general for informatics at the Department of Health, Christine Connelly has had one of the least enviable jobs in public sector IT: defending the much- maligned NHS National Programme for IT (NPfIT). But despite a fiercely critical review in a recent National Audit Office report and at a Public Accounts Committee (PAC) meeting, she told Computer Weekly she stands by the controversial £11.4bn programme.

Christine Connelly was appointed the department's first CIO for health in September 2008, having previously worked as CIO at Cadbury Schweppes. Although she's only been at the helm of the NPfIT for a short period, Connelly says she remains convinced of its importance.

The National Audit Office (NAO) said the programme offers poor value for money, but she strongly disagrees. To get to the point where you receive value for money from your investment, you have to instal the system and let it run for four years, said Connelly: "So it's much too early to say we didn't get value for money."

Another key area of disagreement between Connelly and the NAO is that, despite a massive scaling-back of the programme from its original scope, commensurate cost savings have not been made. "The focus in the report was on what came out of the contracts rather than what went in. It didn't value more functionality. And a one-size-fits-all system is not the way want to go forward," she said.

"With the NPfIT we've focused investment on clinical areas that we think make the most difference to the NHS. So we've put more functionality on the London contract, for example, and let trusts with good systems keep them, rather than ripping them out and putting new ones in for the sake of it. That seemed like a better value proposition than making everybody take a standard system."

However, there's no escaping the project's failures. System implementation has limped along for almost a decade, with many hospitals unlikely to reach completion by 2016. So what went wrong?

"When the programme started out, the approach was to create a standard system that would be the same for everybody. If suppliers had that standard product ready to go there and then it would have rolled out and worked. But it's taken us some time to get to a position where we have a product that is fit for purpose for the NHS. And by the time we got there, the shape of NHS had shifted," said Connelly.

"What we are looking at now is a much softer edge to what the health system was - such as patients being able to exercise choice and go wherever they want for treatment. These things are already there from a policy perspective. And the policies in 2002 compared to 2011 are quite different," she said.

"Moving to a world where we can deploy projects in much smaller chunks and offer trusts choice of which bits of systems they want and how to configure them to local practices is important."

But it's hard to believe the problems were simply caused by a mismatch between policies conceived nine years ago and the NHS as it stands today, and not due to the project's sheer scale.

It's a bit of both, says Connelly. "Given that the product was not there at the start and then expecting that product to get developed and do everything we wanted to do with it - that was probably over-ambitious," she said.

If Connelly were to design the system today, she says she would include tighter contract terms: "When writing a contract that says everything is to hinge on the delivery of a product, it should have a very clear statement that says if the product is not there by an amount of time, then everything else in the contract is then void. That's personally what I would do differently."

Another major criticism of the programme has been its top-down approach to healthcare professionals. But Connelly says the department has done much to improve clinical engagement. "One of the things I thought was disappointing was that the PAC refused to let us bring a clinical witness. I have no idea why [PAC chairwoman Margaret Hodge] made that decision at such short notice, but we think that's a shame."

While Connelly certainly disagrees with much of the recent round of criticism, she does accept the NAO's findings that contracts with key supplier CSC are not working. The NAO it found that two systems would need to be rolled out every month for the contract to be completed by 2016.

"We agree with the NAO's conclusion that the outstanding CSC contracts can't deliver as they stand. We've been in significant negotiations with CSC for some time and we believe there is something that can be constructed out of that to drive value, but all options are being considered. That decision needs to be made across government."

Just three trusts are running Lorenzo release 1.9, the latest version of the CSC patient administration software. "They are early adopter trusts and as early adopter trusts we would expect some levels of difficulty, although not the amount we had," she said.

Connelly flags up the £1.3bn the department has knocked off the total cost of the NPfIT as evidence the department is finding cost savings. But couldn't further money be found by cutting back on the remaining £4.3bn left to spend?

"That £4.3bn does not go out the door until [the systems are] delivered. We don't pay for work in progress," she said. Of that figure, £2.3bn is under contractual commitment. "And that will only be paid if suppliers deliver what they are supposed to deliver."

System interoperability will not be comprised, adds Connelly: "The PAC committee asked, 'will we end up with silos of data we can't join up?' And the answer is that we will not."

It's a familiar lament of many government IT chiefs that projects in the public sector are only discussed in relation to failures. Connelly also believes it's important to remember the successes of NPfIT: "Through the National Programme we've implemented some great shared systems, infrastructure; a shared set of national standards and some fantastic applications - such as the Demographic Service [part of the care record system]. By creating that as a structure in a standard way we now have lots of different systems across the NHS that are using it, not just the systems in the National Programme," she said.

"For example when the Breast Screening Programme does a recall for women to get screened, it doesn't keep a list of all women in the country between 15 and 70, but uses their NHS numbers to go into the Demographic Service to get that information and carry on. And we were only able to set up our 'flu pandemic system last year because the Demographic Service existed."

IT has to be central to NHS policy, says Connelly. This is especially true as the government's controversial changes to the NHS are currently on hold, as it conducts its listening exercise in response to criticism of the reforms.

"There are obviously some possible impacts across our NHS IT strategy, so we're holding back until it's completed. If there are changes then it's important we change too. We can't be separate from all that," she said.

Now is the time for innovation, she believes: "In the 21st century, innovation is often underpinned by IT, so we need to make sure we're having conversations to enable innovation to happen. We also need to do our bit and look at every penny we spend on IT and make sure it is delivering the best value for money that it possibly can."

Although the problems surrounding the NPfIT are likely to continue for some time - particularly with claims CSC has breached its contracts - Connelly says she is optimistic about the future.

"When you look at the world of IT, things move at an exponential pace and when you look at the world today, compared to five years ago, it's dramatically different. Who knows what the next five years will bring? So this is a great place to be at the moment."